{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Estimating network latency and capacity from asymmetric measures\n",
"\n",
"We first load our dependencies and download the provided datasets if required."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"from os import path\n",
"from urllib.request import urlretrieve\n",
"import gzip\n",
"\n",
"import numpy as np\n",
"import pandas as pd\n",
"import statsmodels.api as sm\n",
"import matplotlib.pyplot as plt"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"lan_url = 'http://mescal.imag.fr/membres/arnaud.legrand/teaching/2014/RICM4_EP_ping/liglab2.log.gz'\n",
"lan_file = 'liglab2.log.gz'\n",
"\n",
"if not path.exists(lan_file):\n",
" urlretrieve(lan_url, lan_file)\n",
" \n",
"wan_url = 'http://mescal.imag.fr/membres/arnaud.legrand/teaching/2014/RICM4_EP_ping/stackoverflow.log.gz'\n",
"wan_file = 'stackoverflow.log.gz'\n",
"\n",
"if not path.exists(wan_file):\n",
" urlretrieve(wan_url, wan_file)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## First dataset: Local Area Network conditions\n",
"\n",
"### Data extraction\n",
"\n",
"The first dataset contains ping results. Here is an extract of the provided data."
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"b'[1421761682.052172] 665 bytes from lig-publig.imag.fr (129.88.11.7): icmp_seq=1 ttl=60 time=22.5 ms\\n'\n",
"b'[1421761682.277315] 1373 bytes from lig-publig.imag.fr (129.88.11.7): icmp_seq=1 ttl=60 time=21.2 ms\\n'\n",
"b'[1421761682.502054] 262 bytes from lig-publig.imag.fr (129.88.11.7): icmp_seq=1 ttl=60 time=21.2 ms\\n'\n",
"b'[1421761682.729257] 1107 bytes from lig-publig.imag.fr (129.88.11.7): icmp_seq=1 ttl=60 time=23.3 ms\\n'\n",
"b'[1421761682.934648] 1128 bytes from lig-publig.imag.fr (129.88.11.7): icmp_seq=1 ttl=60 time=1.41 ms\\n'\n"
]
}
],
"source": [
"with gzip.open(lan_file, 'rb') as f:\n",
" for _ in range(5):\n",
" print(next(f))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We create a function to parse these log lines and test it."
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"def parse_ping_line(l):\n",
" assert len(l) > 0\n",
" assert l[0] == '['\n",
" \n",
" timestamp = pd.Timestamp(float(l[1:l.index(']')]), unit='s')\n",
" size = np.nan\n",
" rtt = np.nan\n",
" \n",
" tokens = l.split(' ')\n",
" for i, t in enumerate(tokens):\n",
" if t == 'bytes':\n",
" assert np.isnan(size)\n",
" size = int(tokens[i-1])\n",
" elif t.startswith('time='):\n",
" rtt = float(t[len('time='):])\n",
" \n",
" return [timestamp, size, rtt]\n",
" \n",
"test_line = '[1421761682.052172] 665 bytes from lig-publig.imag.fr (129.88.11.7): icmp_seq=1 ttl=60 time=22.5 ms\\n'\n",
"[timestamp, size, rtt] = parse_ping_line(test_line)\n",
"assert timestamp == pd.Timestamp('2015-01-20 13:48:02.052172')\n",
"assert size == 665\n",
"assert rtt == 22.5"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We can now process our full dataset and check if any data is missing."
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" size rtt\n",
"timestamp \n",
"2015-01-20 13:48:04.770828000 21 NaN\n",
"2015-01-20 13:48:26.146320000 9 NaN\n",
"2015-01-20 13:49:08.647974000 9 NaN\n",
"2015-01-20 13:49:48.500515999 9 NaN\n",
"2015-01-20 13:50:05.331125000 18 NaN\n"
]
}
],
"source": [
"rows = []\n",
"with gzip.open(lan_file, 'rb') as f:\n",
" for l in f:\n",
" try:\n",
" rows.append(parse_ping_line(l.decode()))\n",
" except (AssertionError):\n",
" print('invalid line', l)\n",
"\n",
"df = pd.DataFrame(rows, columns=['timestamp', 'size', 'rtt']).set_index('timestamp').sort_index()\n",
"print(df[df.isnull().any(axis=1)].head())"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The value `time` is missing from some rows. We remove these measurements."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" size | \n",
" rtt | \n",
"
\n",
" \n",
" \n",
" \n",
" | count | \n",
" 44036.000000 | \n",
" 44036.000000 | \n",
"
\n",
" \n",
" | mean | \n",
" 1000.025729 | \n",
" 5.136663 | \n",
"
\n",
" \n",
" | std | \n",
" 573.659666 | \n",
" 12.221312 | \n",
"
\n",
" \n",
" | min | \n",
" 24.000000 | \n",
" 1.000000 | \n",
"
\n",
" \n",
" | 25% | \n",
" 499.000000 | \n",
" 1.270000 | \n",
"
\n",
" \n",
" | 50% | \n",
" 989.000000 | \n",
" 1.440000 | \n",
"
\n",
" \n",
" | 75% | \n",
" 1499.000000 | \n",
" 2.310000 | \n",
"
\n",
" \n",
" | max | \n",
" 2007.000000 | \n",
" 276.000000 | \n",
"
\n",
" \n",
"
\n",
"